Expression of MDM2 mRNA, MDM2, P53 and P16 Proteins in Urothelial Lesions in the View of the WHO 4th Edition Guidelines as a Molecular Insight towards Personalized Medicine

AIM: Here we imposed a multimarker molecular panel composed of P53, MDM2 protein & mRNA & P16 with the identification of sensitive and specific cut offs among the Egyptian urothelial carcinomas bilharzial or not emphasize the pathological and molecular classifications, pathways and prognosis as a privilege for adjuvant therapy. METHODS: Three hundred and ten urothelial lesions were pathologically evaluated and grouped as follows: 50 chronic cystitis as benign, 240 urothelial carcinomas and 20 normal bladder tissue as a control. Immunohistochemistry for MDM Protein, P16 & p53 and In Situ Hybridization for MDM2mRNA were done. RESULTS: MDM2mRNA overexpression correlated with low grade low stage non invasive tumors, while P53 > 40% & p16 < 10% cut offs correlated with high grade high stage invasive carcinomas & bilharzial tumors (P=0.000). CONCLUSION: MDM2mRNA overexpression vs. P53 > 40% & P16 < 10% constitutes a multimarker molecular panel with significant cut offs, proved to distinguish low grade, low stage non invasive urothelial carcinomas (MDM2mRNA overexpression, P53 < 40%, P16 > 10%) from high grade, high stage invasive urothelial carcinomas (with p53 > 40, p16 < 10% & absent MDM2mRNA overexpression). Combined P53 > 40 & p16 < 10%, together with the histopathological features can distinguish in situ urothelial lesions from dysplastic and atypical lesions.


Introduction
Bladder cancer is the 7th most common cancer worldwide [1] and constitutes 30.3% of all cancers in Egypt [2]. Grading of urothelial carcinomas is important in noninvasive disease. Most of the invasive carcinomas are high grade [3].
The 4th edition 2016 WHO guidelines continue to recommend the application of the grading classification of urothelial lesions ISUP 1997 [4] into two major categories as non-invasive group (whether papillary or flat) or invasive group with several advantages, among them are the definite definition of a high grade lesions group which has high risk of progression and can be candidates for adjuvant therapy, with elimination of diagnostic ambiguity particularly grade 2 lesions [3]. Also whereas lowgrade tumours are almost noninvasive (Ta), highgrade tumours are classified based on muscle invasion as Non-Muscle Invasive Bladder cancer (NMIBC; Tis, Ta, T1) or Muscle Invasive Bladder cancer (MIBC; ≥ T2) [5].
Since molecular alterations differ markedly between low and high grade, invasive and not invasive tumours [3], two distinct pathogenic molecular alterations documented: low-grade pathway which involves mutations in FGFR3, PIK3CA, and inactivating KDM6A mutations, whereas high-grade muscle-invasive tumours pathway shows TP53 and RB1 alterations [6]. Moreover, with the use of whole genome mRNA expression profiling, three intrinsic unique molecular subtypes of muscle invasive bladder carcinomas documented, strikingly recapitulating molecular subtypes of breast cancer as basal type, luminal type and p53-like muscle-invasive type tumours. The p53-like tumours are resistant to neoadjuvant chemotherapy. Thus, all chemoresistant tumours adopted a p53-like phenotype after therapy [7][8].
The p53 protein is encoded by the TP53 tumor-suppressor gene which located at 17p13.1 [9]. It inhibits cell-cycle progression at the G1-S transition. Altered p53 expression increases progressively from normal urothelium to in situ urothelial carcinomas (flat or papillary) to not muscle-invasive, to muscleinvasive disease with metastatic lymph nodes [8,10]. P53 used for urothelial carcinoma stratification [8,11] however alone is not sufficient to suggest p53 alterations and can't be used as a prognostic marker in urothelial carcinomas or chemotherapeutic response stratification [8,12]. Therefore, a combination of p53 with other molecular markers improves risk stratification [8,12].
Human MDM2 gene located at 12q13-14 [13]. It contains a p53 binding domain. MDM2 protein is an oncoprotein has a negative regulating effect on p53 [14]. MDM2 gene amplification is infrequent in bladder cancer despite elevated MDM2 protein levels [15]. MDM2 expression correlates with tumour grade and recurrence in superficial bladder cancer. Acquisition of MDM2 gene expression significantly associated with high tumour grade [16,17]. Amplification of MDM2 may increase sensitivity to MDM2 antagonists [18]. In addition, the absence of MDM2mRNA was reported in bilharzial tumours indicating their aggressiveness and poor prognosis [18].
P16 is an early marker for malignant transformation in human cancers [19,20]. P16 is a TP53-related kinase that controls cell cycle progression. P16 gene locates at 9p21, which is a major site for deletions in bladder cancer. The p16 gene abnormalities are predominant in schistosomal SQCC than conventional TCC. P16 and TP53 show mutually complementary role in the pathogenesis of bladder cancer [21,22]. Loss of p16 expression is significantly associated with high-grade tumours and reduced progression-free survival [21], particularly in early stage bladder cancers [22].
Moreover, mutations of genes associated with cell cycle control were detected in schistosomal bladder cancers [23]. Early deletion of P16 gene occurs in schistosomal TCC than schistosomal SCC [24]. This is due to chromosomal instability induced by bilharzial irritation [23][24][25][26]. Molecular subtyping of bladder cancer is a tool for personalized medicine [8]. No molecular biomarkers are widely used for clinical outcome prediction [21].
Our study aimed at establishing an applicable multimarker molecular panel of MDM2 (mRNA & Protein), P16 and p53, with sensitive and specific cut offs, to stratify Egyptian urothelial carcinomas according to their molecular pathways within the context of WHO 2016 grade and stage classification, and to predict tumor progression & prognosis, particularly for high grade carcinomas, as a privilege for adjuvant therapy.

Patients and Methods
The study held on 310 urothelial lesions obtained from archival paraffin blocks at pathology department TBRI (2012-2016) and grouped as follows: 50 chronic cystitis as benign, 240 urothelial carcinomas and 20 normal bladder tissue as a control.

In Situ Hybridization for MDM2mRNA
Paraffin-embedded sections were deparaffinized and treated with prehybridization mixture. ISH was performed overnight. Human MDM2 cDNA probe kindly provided by Bruno Voss (Professional Associations' Research Institute for Occupational Medicine BGFA, Ruhr-University, Bochum, Germany) was used. The reaction signals were amplified by Tyramide signal amplification (TSA) kit (Invitrogen, Grand Island, NY). Counterstaining was done. Slides washed in PBS buffer and kept moist with glycerol.

Assessment of MDM2 mRNA in Situ Hybridization, MDM2 protein, p16 & p53 immunostaining
 P16 was considered positive when at least one atypical cell with strong nuclear expression with or without cytoplasmic positivity, independently of the percentage of positive cells. Cases with weak or absent nuclear expression were considered negative [27].

Statistical analysis
Statistical evaluation was performed with SPSS (version 20, IBM, Chicago, Il, USA). The correlation between expressions and clinicopathological parameters was assessed using Spearman's correlation test. Differences between proportions studied using Chi square test were deemed significant at the level of p < 0.05. Sensitivity, specificity, false positive & negative rates were calculated.

Results
MDM2 protein was expressed in 47.1% of malignant cases. All control and benign cases were negative. Similarly, MDM2mRNA was expressed in all malignant cases with negative expression in both benign and control cases (P<0.000) ( Table 1 and P53 showed positive expression in 50% of benign cases and all malignant cases (P < 0.01). Surprisingly most of the non-neoplastic cases showed rather a variable expression up to 40%. However, none expressed P53 > 40% in comparison with urothelial carcinomas (P < 0.001).
MDM2 mRNA score was significantly inversely correlated with the MDM2 protein expression, P53 expression & score, as well as P16 score (p < 0.001). In contrast, MDM2 protein expression was significantly directly correlated with P53 expression & score, as well as P16 score. Moreover, P53 expression & score were directly correlated with each other, with the P16 score and with MDM2 protein expression, however inversely correlated with MDM2 mRNA score (Tables 2 and 3).

Discussion
The challenging task in molecular pathology analysis is to establish the clinical relevance of molecular types beyond the histopathologic appearance [31]. Bladder cancer shouldn't be managed only according the clinical or pathological features. The molecular approach in combination with the clinicopathological features is mandatory. Nevertheless, a routine use of a well-established molecular approach (as in breast cancer) still not identified or recommended [32]. And due to tumour heterogeneity, no single marker can reflect the tumour biology. Therefore, combined biomarkers panel improves predictive and prognostic accuracy in order to get a personalized treatment approach [33].
Here we imposed a multimarker molecular panel composed of P53, MDM2 protein & mRNA & P16 with the identification of sensitive and specific cut offs among the Egyptian urothelial carcinomas bilharzial or not. The target is to emphasize their pathological classification according to the WHO 4th edition into low vs. high grade particularly in ambiguous cases and to facilitate prediction of potential course, pathways & prognosis.
In our study, all non-neoplastic cases showed neither MDM2 protein nor MDM2mRNA expression. Similarly, El-Abd et al., 2008 [18] showed the absence of MDM2mRNA in controls and benign cases. However, our study showed that the overexpression of MDM2 mRNA significantly indicated low-grade low stage and rather non invasive tumours with sensitivity (87.5 %) and high specificity (98.4 %), in contrast to high-grade, high stage invasive tumours. This came similar to Schlott et al., 2004 [34] in which MDM2mRNA was significantly 5-folds lower in advanced high grade, high stage urothelial carcinomas. Moreover, we showed that MDM2 mRNA score was inversely correlated with the MDM2 protein expression. Similarly, it was mentioned that MDM2 gene amplification is infrequent in bladder cancer despite elevated MDM2 protein levels [34]. It is amplified in 10% of urothelial carcinomas [15]. Furthermore, we that MDM2 mRNA score was significantly inversely correlated with the MDM2 protein expression, P53 expression & score, as well as P16 score. This may be due to that MDM2 amplification doesn't occur with p53 mutations within the same tumour, indicating that carcinogenesis results from MDM2 amplification alone [34]. This also came along with Pfister et al., 2000 [35] who stated that tumours overexpressing MDM2 but not p53 are rarely of high grade. In the same context, Uchida et al., 2002 [36] showed that co-expression of p53 and MDM2 with MDM2 overexpression is associated with favourable prognosis in invasive carcinomas. This is due to loss of MDM2 inhibition and/or DNA damage resulting in increase of wild-type of p53 at a level not sufficient for immunohistochemistry detection [35]. However if p53 gene become mutated, tumor will show higher grade, worse prognosis, higher recurrence rate and shorter progression time & survival in contrast to wild type non-mutant P53 [22, [22,38]. Mutant TP53 prolongs survival of cells with established genetic defects, allowing them to become more unstable and aggressive [22].
Moreover, surprisingly 50% of benign cases showed rather a variable expression of up to 40%, however none expressed P53>40%. Thus P53>40 cut off together with the cellular morphology can identify and distinguish the in situ urothelial lesions versus otherwise urothelial atypia or dysplasia. This came along with Cheng et al., 2014 [22] since the p53 nuclear expression is not always indicative of TP53 mutations, and not all TP53 mutations result in protein accumulation [22].
Immunohistochemistry relies on the accumulation of p53 protein due to prolonged half-life of cells. The half-life of wild type p53 is estimated between 20 and 30 min, whereas mutation there is decreased degradation of mutant p53 which has a longer half life up to 24 hours [22,39]. A TP35 mutation is a late event in carcinogenesis leads to loss of the remaining wild type allele and inactivation of growth control function [22,39] resulting in an altered protein resistant to degradation, and shows nuclear accumulation, altered DNA repair, cancer development and progression [31,41].
The low p16 expression is associated with tumorigenesis [23,42,43]. Our study interestingly showed p16<10% expression in benign cases (P<0.001). Similarly, the possibility of P16 gene deletion in some benign and control cases was reported [25]. Nevertheless, the absence of p16 was also mentioned in benign urothelium [44]. Therefore, combined P16<10% and P53>40% together with cellular morphology can distinguish in situ urothelial lesions versus otherwise atypia or dysplasia. P16<10% cut off was significant in high grade, high stage, invasive tumours with (78.50%) sensitivity and (42.40%) specificity. Emphasis on the value of low p16 in the early diagnosis and prognosis of early stage bladder cancer as it indicates poor progression and recurrence free survival rates were reported [21,23,45].
P16 expression in bilharzial SQCC in comparison to non-bilharzial TCC was mentioned [22] and showed similarity to our results however with no significance. Additionally, we showed that bilharziasis significantly directly correlated with MDM2 protein & P53 positivity, P53 & P16 scoring, stage, grade & invasion and exhibited predominant P16<10% in tumours, while inversely correlated with MDM2 mRNA score, indicating aggressiveness and poor prognosis [18,23].